3500 Swap Finally Underway, a few questions

[jmgtp]

What's left that I can think of: torque the rear exhaust manifold to spec, reinstall the crossover pipe, snap the axles into the trans, torque the hubs down, fill the trans oil, install brakes, tie rod ends, alignment, serpentine belt, power steering lines, coolant, needs an airbox/filter. Put the hood back on. Tuning.

I think I literally have only 1 weld left to make on the exhaust.

If I had all the remaining parts on hand and the physical ability, it is a Saturdays worth of work to get it in a driveable state. I guess I can be a little more optimistic that I can finish it in time since it really isnt much work left. If I could get the mechanical stuff done in the next 2 weeks that should still leave me some time to get a decent tune that will pass the sniffer test. Even if I can't find the dogbone bracket for the drivers side I'm sure it could be driven gently with just the passenger side one, also keep in mind that I used the pushrod lower mount that is beneath the balancer and the DOHC rear mount so it is already a little stiffer than a normal pushrod engine would be. I didn't use the DOHC front mount though, thats what lets the dohc get away with only 1 dogbone on the drivers side. I wanted to take out the lowering springs in the front and put the stock height springs back but that can wait.

[Jonpro03]

I'm running just one engine mount, although it's on the other side. Should be ok.

[jmgtp]

Was feeling a bit better today, took my pain meds, threw the back brace on and went to work on it! I finished the exhaust.

Then of course, I started it up! It runs smooth and quiet. BUT it does not run for long. About 10 seconds and then it stalls dead. It is idling far too high as well, about 2500 rpm. So, I either have a vacuum leak, a way out of whack IAC or possibly since there is no coolant in it the sensor thinks it is absolutely freezing and the ecm is intentionally idling that high.

There are 2 codes, 22 and 43. So, 22 is the TPS. I will check the wiring but that is not wiring that I touched so I'm not sure what is going on with it. While the engine is running if I open the throttle at all it will die instantly so there is for sure a TPS issue.

Code 43 is I believe the code indicating an issue with the ECMs ability to manipulate the spark timing. What is the typical cause of that?

I also don't know why the car will run 10 seconds and then die instantly. At first I thought maybe the fuel pump was cutting out because it wasn't detecting oil pressure (the dash gauge has good pressure). So I jumpered the fuel pump test lead to make it run constantly and it still stalls after 10 seconds. So it is not the fuel pump.

This weekend I should be able to have the lower radiator hose installed (not sure if I got right one), and I got a belt for it too then I can fill with water and see if that makes a difference. I guess first thing to check would be correcting whatever is causing the 2 codes.

At least its some progress. I'd like to thank percocet and valium for getting me there (and no I don't work on it when under the influence of either).

[jmgtp]

Fixed the TPS sensor and it runs without stalling, but still has code 22 even after a computer reset. Haven't looked into code 43 yet.
I now also have code 34 which is a MAP sensor problem. I checked the wiring with an ohm meter and it all checks out good. I converted to the new small style map, but I don't see that being an issue.

3 other problems:

1.) Code 15, Coolant temp sensor. I posted about this in pushrod general. I think its because I used the LX9 CTS which probably doesnt output the same values as the old style sensors. Problem is that the old sensor is much too large to thread into the LX9 cylinder head. I don't know what the fix for this is.

2.) I put the serpentine belt on. The alterantor isn't charging. It tested good at the store. The original LQ1 alternator had 2 wires on the connector, one 12v and one ground. But, the 3x00 alternator that I am using now only has a 1 wire connector which I believe was just 12v. So the ground wire in my harness is not connected to anything, could this be the problem?

3.) I cannot get TunerPro to link up to my car. I have a new laptop now and I just can't get it to work. I'll post in the computer section about this.

[CNCguy]

Did you ever find the bracket? I was looking for something today and found one at the shop. If you still need it, it's yours for shipping.

[jmgtp]

I never found it so if you have one I will definitely take it.

Just to be sure, it is the bracket that would bolt to the cylinder head on the drivers side of a 3x00 W body car? If you have a pic I can confirm whether or not it looks right.

Thanks,
John

[CNCguy]

That's the bracket and I will try to get a pic tomorrow. It's from a '96 Lumina and I believe the links are there as well if you want them. I do not have the pieces that bolt to the core support.

[Jonpro03]

what passanger side dogbone mount are you using? I don't have one installed but I'd like to. What did you get it off of?

[WrathOfSocrus]

The lower bolt bosses on the block where they meet the passenger mount on 3x00 W-body and minivan are shorter than the 3500. If u take one of those mounts you can see a casting seam which is roughly the amount needed to be removed. I just cut it off above the seam with an angle grinder and ground it down a bit from there until it looked like it fit correctly. I'm sure you could break out a caliper and measure it to the thousandths if you feel it is needed. The attachment I put up a while ago shows the bracket and if u look at the bottom portion you can see the shiny part where it was cut.

[jmgtp]

I want to say the passenger side bracket that I used was from an early 3100 or a 3.1 w body car.
3 of the bolts lined up, but it did require some cutting/grinding to fit. I mainly wanted that one because it lets me use my stock a/c compressor. The bracket that wrathofsocrus posted is not the one I used. That one is meant for the later style a/c compressor.

I got freescan to work and tunerpro too. It needed new drivers for the usb/serial port adapter.

I fixed code 34 (bad 5v ref on the map sensor) and code 15 (bad ground on the CTS).
Code 22 is resolved as well. Code 43 though - the knock sensor, the wire tests good from the connector to the ecm so I dont think its wiring. It is the same KS pulled from my dohc.

Now, the engine runs and will run continuously but it is running horribly. It feels like its running on 4 cylinders or something. The only code I still have is 43 (knock sensor) and I don't know if that is the cause of it running poorly or the result.

I double checked that each ignition wire was going to the right coil - they are. I did reuse the LX9 wires but had to crimp new ends on for the older style DIS coils. So I'm thinking either I have a bad wire, bad coil or bad ICM. Good thing is I have spare ICM/coils to throw on. Bad thing is I dont have spare wires, so I'm going to suck it up and buy a new set.

Also, it seems to run better on the stock 3.4 dohc bin than it does on the 3500 bin that I have, but neither runs good at all. When you rev it, it sounds like it is fighting to do so. The exhaust smells pretty bad and it did backfire out of the intake at least once on a failed start attempt. Gotta be ignition problem right?

Other than a short to ground what else could cause a code 43?

[bszopi]

I've heard different opinions on this, but you should be matching the KS to the memcal, as the memcal houses the KS filter circuit.

What do your logs look like? Post them up for people to check out.

Droid via Tapatalk

[jmgtp]

I've heard the same, but regardless this is the same memcal and knock sensor from the old motor which didn't trip code 43.
I didn't datalog it yet because it is running so poorly that I don't run it for more than 10 seconds or so because it just sounds like it will break.
There has to be a huge problem somewhere, I just havent found it. The engine really feels as if I have the spark plug wires crossed or have a dead wire. I double checked the wires to the coil location and it looks right. I'm going to end up pulling each wire and checking it with an ohm meter. Then I'll replace the ICM and coils with a spare that I have. Hopefully that cures it.

I might take a break from the electrical and finish the mechanical, that is put the brakes and wheels back on, so at least I can get this thing to the front of the garage so when I run it the fumes go outside. Once I have it at the door I'll feel comfortable running it longer and I'll try to datalog it, at least that way it won't poison me. Freescan seems to hang up when the motor is running but works great with the motor not running. TunerPro is running at a stupid slow refresh rate and I don't know how to make it go faster.

[Jonpro03]

You have to use the 3500 KS. The circuitry is internal to the sensor and it's specifically tuned to the 3500.

The knock sensor is tuned to a specify frequency, like a tuning fork. When this frequency is applied to the sensor (through its connection to the engine), a piezoelectric crystal inside the sensor generates a small voltage (~1 volt), much like a microphone. As an example, some Corvette knock sensors (GM PN 1997562, 1997699, OR Standard Motor Products KS45, KS46, KS49, or KS117) have a design frequency of 5200 hertz, and they produce a signal between 4800 rpm and 5600 rpm.

My source http://www.megamanual.com/ms2/knock.htm

[jmgtp]

I'm not using a 3500 computer though. I'm using my OBD I computer. I have a .bin that was 'tuned' for a 3500 engine that I tried and it didn't run well, so I went back to the 3.4dohc .bin and it runs slightly better (still not well).

I don't think the obd I computer can use the LX9 knock sensors. It was suggested (in another thread) that I check to make sure the KS is grounding correctly and that is something that hadn't crossed my mind. If I put too much teflon tape or if there was excessive corrossion where it threads into the block it might not be grounding correctly. Edit: Just to be clear in what I am saying, the LX9 knock sensor is a completely different kind of knock sensor compared to what the older ones. I cannot use the 3500 KS because it is not compatble with the ecm I am using.

I'm hoping to be able to resolve the knock sensor issue tonight, whether it be a bad ground, sensor or wiring.

I really wanted to reuse the LX9 ignition wires since they are OEM which is typically higher quality than aftermarket and they looked to be in excellent shape. But this meant crimping the correct terminal ends onto the wires for the older style DIS which my comptuer uses. I could have a bad wire(s) causing my problem. I'm going to eat teh cost of a new set for a 3400 just to be sure that wires aren't the issue. I will also install a spare ICM and coils that I have to ensure that they are not the cause.

Before I do any of that I need to put the brakes/wheels on and put the back of the car near the garage door. The car is too far in the garage right now and when I run the motor the exhaust comes in rather than goes out. I know brakes and wheels sounds like not a lot of work but I'm still suffering back probs and on pain meds so doing putting them on and getting the car on the ground will probably wipe me out.

[Jonpro03]

http://imgur.com/NrCjX.png

If I were you, I would be wiring in the 3500 knock sensors. Your 3.4 ECM will be able to understand them just fine.

The light blue wire goes to the 1 wire that your ecm has run to the old KS. The dark blue is a low reference that should be run to a black wire coming out of the ECM harness or you could ground it to the block and it should still work. You could also shield it if you want to. I don't think your 3.4 KS was shielded though. Just keep the lines away from the alternator and coils/plugwires and you should be ok.